All Stats Formulas Flashcards
Current up till term 2 holiday
Standard deviation
√{ Σx² - n(x̄)² } / n
Coded mean
{ Σ(x ± a) / n } ∓ a
Coded standard deviation
√ { Σ(x ± a)²/ n } - { [Σ(x ± a) / n]² }
Combined mean
[Σx + Σy] / [nₓ + nᵧ]
Combined standard deviation
√ { [Σx² + Σy²] / [nₓ + nᵧ] } - { [Σx + Σy / nₓ + nᵧ]² }
Applied mean
1/a{ Σ(ax ± b) / n } ∓ b
Applied standard deviation
1/a √ { Σ(ax ± b)²/ n } - { [Σ(ax ± b) / n]² }
Objects in a circle (without repeats)
(n -1)!
Independent events
P(A & B) = P(A ∩ B) = P(A) x P(B)
Conditional probability
P(A|B) = { P(A ∩ B) } / P(B)
Outlier
1.5x greater the UQ or 1.5x lesser than LQ
Skewness
skew refers to tail so positive skew has tail towards positive and negative skew has tail towards negative
Mutually exclusive events
P(A ∩ B) = 0 , A and B can not occur together
Probability summation
Σ P(X = x) = 1
Expectation
E(X) = µ = ∑ xᵢPᵢ
Variance of a random variable (2)
Var(X) = E(X - µ)² = ∑ (xᵢ - µ)² x P(xᵢ)
Var(X) = E(X²) - [E(X)]²
Conditions/Notation of Binomial Distribution (5)
X ~ B (n, p)
- Only 2 possible outcomes (discrete) usually called successes (p) or failures (q), [successes (p), outcome we are interested]
- There are a fixed number of trials (n)
- Each trial must be independent of the other trials
- The probability of success (p) is fixed at each trial
Binomial distribution (4)
P(X = x) = (n x)pˣ(1-p)ⁿ⁻ˣ = (n x)pˣqⁿ⁻ˣ
E(X) = np
Var(X) = npq
Sd(X) = √Var(X) = √npq
Mode of Binomial Distribution
X value with the highest probability
Conditions/Notation of Geometric Distribution
X ~ Geo (p)
- Only 2 possible outcomes (discrete) usually called successes (p) or failures (q), [successes (p), outcome we are interested]
- The repeated trials can be infinite
- Each trial must be independent of the other trials
- The probability of success (p) is fixed at each trial
Geometric distribution (4)
P(X = x) = p(1 - p)ˣ⁻¹ = pqˣ⁻¹
E(X) = 1/p
Var(X) = q/p²
Sd(X) = √Var(X) = √q/p²
Geometric Distributional Inequalities
Fewer than:
P(X > x) = (1 - p)ˣ = qˣ
At least:
P(X ≤ x) = 1 - (1 - p)ˣ = 1 - qˣ
Mode of Geometric distribution
P(X = 1) has the greatest probability in all geometric distributions
Standard normal distribution
X ~ N (µ, σ²) –> Z ~ N (0, 1)
Standardising normal distribution
X ~ N (µ, σ²) where Z = [ x - µ ] / σ
Normal approximation to binomial distribution
If X ~ B (n, p) where np > 5 and nq > 5 then
X ~ B (n, p) –> X ~ N (µ, σ²)